The clinical phenotype associated with myositis-specific and associated autoantibodies: a meta-analysis revisiting the so-called antisynthetase syndrome

Objective: To describe the clinical spectrum associated with aminoacyl-transfer RNA synthetase (ARS) autoantibodies in patients with idiopathic inflammatory myositis defined according to Peter and Bohan's criteria.

Methods: Cohort studies were selected from MEDLINE and Embase up to August 2013. Two investigators independently extracted data on study design, patient characteristics, and clinical features (interstitial lung disease [ILD], fever, mechanic's hands [MH], Raynaud's phenomenon [RPh], arthralgia, sclerodactyly, cancer and dermatomyositis-specific rash) according to the presence of myositis-specific (anti-aminoacyl-transfer RNA synthetase [ARS], anti-signal recognition particle [anti-SRP] and anti-Mi2) and myositis-associated (anti-PM/Scl, anti-U1-RNP and anti-Ku) autoantibodies.

Results: 27 studies (3487 patients) were included in the meta-analysis. Arthralgia (75%, CI 67-81) and ILD (69%, CI 63-74) were the most prevalent clinical signs associated with anti-ARS autoantibodies. Anti-Mi2 and anti-SRP autoantibodies were associated with few extramuscular signs. ARS autoantibodies were identified in 13% of patients with cancer-associated myositis (5-25). Patients with non-anti-Jo1 ARS had greater odds of presenting fever (RR 0.63, CI 0.52-0.90) and ILD (RR 0.87, CI 0.81-0.93) compared to those with anti-Jo1 autoantibodies. The frequencies of myositis (RR 1.60, CI 1.38-1.85), arthralgia (RR 1.52, CI 1.32-1.76) and MH (RR 1.47, CI 1.11-1.94) were almost 50% higher in patients with anti-Jo1 compared to non-anti-Jo1 ARS autoantibodies. Patients with anti-PM/Scl differed from those with anti-ARS autoantibodies by a greater prevalence of RPh (RR 0.70, CI 0.53-0.94) and sclerodactyly (RR 0.47, CI 0.25-0.89). ILD was less frequent in patients with anti-U1-RNP autoantibodies (RR 3.35, CI 1.07-10.43). No difference was observed between anti-ARS and myositis-associated autoantibodies for other outcomes.

Conclusions: The presence of anti-ARS autoantibodies delimits a heterogeneous subset of patients with a high prevalence of myositis, MH, arthralgia in anti-Jo1 patients, and RPh and fever in non-anti-Jo1 patients. The clinical signs of populations positive for anti-PM/Scl and anti-ARS autoantibodies largely overlap, especially with regard to ILD, challenging the clinical delimitation of the antisynthetase syndrome.